I
ALFRED C. OLSON‘ California Research Corp., Richmond, Calif.
Alkylation of Aromatics . . . with 7-Alkenes Arylalkane mixtures from the same olefin vary in composition when made in the presence of hydrofluoric acid, sulfuric acid, or aluminum chloride
A L R Y L B E N Z E N E S U L P O N A ~ E S are
widely used by leading soap manufacturers in their commercial detergent products. These alkylbenzenesulfonates are prepared by alkylating benzene with olefins followed by sulfonation and neutralization. The position of the phenyl group on the alkyl chain is an important factor in determining the surface active properties of these sulfonates (3, 6, 7). For this reason, studies concerned with isomerization during alkylation and the final attachment of the phenyl group to the alkyl chain are of considerable interest. It has been reported that isomerization occurs during alkylation of benzene with I-heptene in the presence of sulfuric acid (2) and 1-dodecene in the presence of “various technical catalysts” (74) with the formation of mixtures of all the secondary phenylalkanes possible without carbon skeleton rearrangement. This is contrary to earlier reports of no isomerization and formation of only 2-arylalkane (8, 70) and partial isomerization to give mainly the 2- and 3isomer ( 7 ) . The work reported here is part of a study of the aluminum chloride, sulfuric acid, and hydrofluoric acid catalyzed alkylations of aromatics with 1alkenes. For all three catalysts the monoalkylation products were composed of mixtures of secondary arylalkanes. The manner in which they were formed in sulfuric and hydrofluoric acid catalysis Present address, Western Regional Research Laboratory, U. S. Department of Agriculture, Albany, Calif.
was different from that in aluminum chloride catalysis.
coal as catalyst. Where necessary, the crude phenyldodecanes were passed through a silica gel column to remove oxygenated compounds (Table I). They were better than 9870 pure as established by mass spectrometric and gas-liquid chromatographic analyses. Alkylations and Isomerizations. Both odd and even numbered 1-alkenes (Humphrey Wilkinson and Co., Baltimore, Md.), analyzed by infrared and gas-liquid chromatography, were better than 9601, pure. The principal impurities in the even numbered compounds were isomeric alkenes. Odd numbered 1-alkenes were prepared from the appropriate alkyl magnesium bromide and allyl bromide. The amounts of reactants used in the alkylations of benzene with 1-dodecene are shown in Table 11. I n all the alkylations with hydrofluoric acid a mole ratio of alkene-aromatic-hydrofluoric acid
Experimental
Isomeric Phenyldodecanes for Analytical Standardization. 1-, 2-, 3-, 4-,5 - , and 6-Phenyldodecane were prepared by standard procedures. 1-Phenyldodecane was prepared by the aluminum chloride catalyzed acylation of benzene with dodecoyl chloride. The ketone from this reaction was reduced to 1phenyldodecane by a modified Wolff Kishner reduction. 2-, 3-, 4-, 5-, and 6-Phenyldodecane were prepared from the appropriate alkylarylketone and alkyl magnesium bromide (Grignard reaction). I n each case, the tertiary alcohol was dehydrated over copper sulfate or potassium bisulfate and the olefin formed was hydrogenated to the phenyldodecane with platinum on char-
Table I.
Properties of Pure Phenyldodecanes Boiline Point Converted,a
Phenyldodecane 123456-
Mm.Hg
mm.Hg
n %o
Relative Retention Timesb
5 5 15 10 10 5
328 312 311 314 304 304
1.4824 1.4817 1.4818 1.4810 1.4812 1.4813
3.8 2.8 2.4 2.1 1.9 1.9
Recorded, O
c.
166.1-166.7 154.0-154.4 172.2-176.1 164 -166 155 -161 144.4-148.3
O
C., 760
a Used midpoints of the boiling point ranges and converted to C., at 760 mm. of Hg using published nomographs (6). Gas-liquid chromatography where I-hexadecane = 1 .O.
VOL. 52, NO. 10
OCTOBER 1960
833
Table II. Analyses Show Mixtures of Isomeric Phenyidodecanes Formed in Alkylation of Benzene with 1 -Dodecene
Alkylation conditions Catalyst Moles catalyst Moles l-dodecene Moles benzene Product, mole yo 1-Dodecene converted to phenyldodecane Didodecylbenzeneb Phenyldodecane isomer distribution, wt.%c Phenyldodecane 1-
1
2
HF 50
AlCla"
H1S04
5 ml. 98y0
5 50
2 10
92
68
72
5
21
(9)
0
0 32 22 16 15 15
41 20 13 13 13
20 17 16 23
23456-
0.1
3
24
0.15 1.5
0
Added 0.1 g. of water to the reaction. tentatively from jnfra.red and mass spectrometer analysis and boiling point d a t a in Experiments 1 and 2 . I n Experiment 3 t h e didodecyl value was obtained from the weight of the material boiling above the phenyldodecanes. Based on gasliquid chromatography analysis. Although no separation of the 5- and 6- isomer was possible by this technique, mass spectrometer analysis showed both isomers Eresent in about equal amounts in all three experiments. a
* Identified
of 1 : 10 : 10-20 was used. I n a typical reaction with this catalyst, aromatic mixed with alkene was added rapidly to the anhydrous acid in a cooled, stainless-steel or polyethylene reactor. The mixture was stirred during the addition and the reaction temperature was kept at 16' + 3' C. The aluminum chloride alkylations were run at 30' and 53' C. by adding 1-dodecene in one half of the stated amount of benzene to the remainder of the benzene containing the aluminum chloride and a promotional amount of water. The reaction was stirred for 15 minutes after the addition. Sulfuric acid alkylations were conducted at 0' to 10' C. by adding 1dodecene in one half the benzene to the remainder of the benzene plus the sulfuric acid. The reaction was stirred for 2 hours after the addition a t 5' to 10' I n all cases products were isolated by separating the catalysts from the reaction mixtures and washing the organic phases free of catalyst with water and aqueous caustic. After removing the unreacted aromatic the crude products were distilled and the material
c.
834
boiling in the range of the arylalkane was analyzed using gas-liquid chromatography and high mass spectrometry. Attempted isomerizations of 1-, 2-, and 6-phenyldodecane in the presence of hydrofluoric acid were carried out by stirring 1.0 gram of the pure isomers separately with benzene and hydrofluoric acid a t 16' =t3' C. Mole ratio of reactants was 1 : 10 : 50 phenyldodecane - benzene - hydrofluoric acid. Contact time was 2 hours. The entire reaction mixture was decomposed in ice water, the organic phase washed free of catalyst, stripped of benzene, and the residue was analyzed directly by gas-liquid chromatography. Product Analyses. By using rhe pure isomeric phenyldodecanes, it was found that 1-, 2-, 3-, and 4-phenyldodecane could be separated from each other and from the 5- and 6- isomer on a gas-liquid chromatographic column (Table I). The column was filled with 35 grams of material consisting of 92 penetration asphalt from a Middle East crude oil and 40- to 60-mesh JohnsManville C-22 firebrick in a ratio of 40 to 100. The flow rate of helium at the column exit was 35 ml. per minute as measured by a soap bubble flowmeter. Column temperatures were 240', 270°, or 320° C., depending on the molecular weight of thc samples being analyzed. As shown by analysis of synthetic mixtures, the amounts of secondary phenylalkanes were within of the actual amount present in the range 10 to 50%. In the range 0 to 1Oyo>they were within =i=20%. Based on the analysis of a syntheric mixture, less than 0.0570 of 1-phenyldodecane could have been detected if it had been present in any of the reaction mixtures. The phenyldodecane mixtures were also analyzed using the high mass spectrometer. Fragmentation patterns of each isomeric phenyldodecane are sufficiently different from each other to permit their identification in mixtures ( 7 7, 74). Analysis of several of the alkylation reaction mixtures by this technique confirmed the results obtained using gas-liquid chromatography and in addition showed the presence of roughly equal amounts of the 5- and 6- isomer. Analyses were done on a Consolidated Electrodynamics Corporation Model 103B mass spectrometer equipped with a gallium fritted disk-heated inlet. Input temperatures were about 320' C., and inlet sample pressure was about 100 microns. Infrared analyses were done on either a Beckman IR-4 or Perkin Elmer 21 recording spectrophotometer using sodium chloride optics and cells. Mixtures of isomeric secondary phenyldodecanes were produced when benzene was alkylated with I-dodecene with either aluminum chloride, sulfuric
I N D U S T R I A L AND E N G I N E E R I N G C H E M I S T R Y
acid, or hydrofluoric acid as catalyst. The compositions of the mixtures are shown in Table 11. As expected nn 1-phenyldodecane was found in any of the products. The isomerization that occurred when hydrofluoric acid was used as the catalyst produced about an equal mixture of 2-, 3-, 4-, 5-, and 6phenyldodecane. Less isomerization occurred when either aluminum chloride or sulfuric acid was used as the catalyst. In these cases 32 and 417,, respectively, of 2-phenyldodecane were formed together with decreasing amounts of the other isomers. Monoalkylation product corresponded to 68, 72, and 927, yield of the phenyldodecanes for the aluminum chloride, sulfuric acid? and hydrofluoric acid catalyzed reactions. The principal side reaction in the case of the aluminum chloride and hydrofluoric acid reactions was the formation of a high boiling fraction identified as didodecylbenzene bv boiling point, mass spectrometer, and infrared analysis. With the sulfuric acid catalyzed reaction some didodecylbenzene probably formed together with a considerable amount of dodecylsulfate. Hydrofluoric Acid Alkylation. nPhenylalkanes from the hydrofluoric acid catalyzed alkylation of benzene with 1alkenes from 1-hexene to 1-octadeceiie were ana.l\rzed by ga2-liquid chromatography. The analyses showed that in every case mixrures of phenylalkanes were formed (Table 111). Structures were assigned to the chromatographic pelks from these analyses following the pattern observed with the phenyldodecanes. Thus, those isomers with the phenyl group farthest from the ends of the alkyl chains were assigned the shortest column retention times in each series. The logarithm of the column retention times of each positional isomer-e.g., 2-phenylalkanes and 3-phenylalkanesgave a straight line when plotted against carbon number. This is consistent with other published work which shows such a relationship between the members of a homologous series of compounds (9, 73). The isomerization occurs prior to final attachment of the phenyl group to the alkyl chain with this catalyst since 2- and 6-phenyldodecane were recovered unchanged after prolonged treatment under hydrofluoric acid alkylation conditions. 1-Phenyldodecane was also recovered unchanged after treatment under hydrofluoric acid alkylation conditions. Thus, if any of this isomer had formed during the alkylation of benzene with 1-dodecene, it would not have isomerized to another isomer. The monoalkylate isolated from an incomplete hydrofluoric acid alkylation had the same mixture of isomers as that from a reaction in which all the olefin was consumed. I n this exper-
AROMATIC ALKYLATION
Table 111. Analyses Show Alkylation of Benzene with 1-Alkenes Gives Mixtures of Secondary Phenylalkanes (Olefin-benzene-HF mole ratio, 1:10:10) Phenylalkane Isomers, % ComMonoalkylbined benzene 5-, 6-, 7-, Source 2348-, 91-Hexene 37 63 1-Octene 33 32 35 1-Decene 23 22 17 38 1-Hendecene 20 19 16 45 1-Dodecene 20 17 16 47 I-Tridecene 17 15 12 56 1-Tetradecene 18 14 14 54 1-Pentadecene 18 14 14 54 1-Hexadecene 24 14 10 52 1-Heptadecene 23 14 9 54 I-Octadecene 31 17 9 44
..
..
..
iment, water was added to the alkylation to quench it before completion so that the phenyldodecane isolated corresponded to 19% of the original olefin. Dodecyl fluoride and dodecene were also recovered in amounts equivalent to 19 and 52Yo of the original olefin. From infrared analysis very little, if any, isomerization of the 1-dodecene occurred during the partial reaction, even though the distribution of secondary phenyldodecanes appeared to be random.
The position of the fluorine in the dodecyl fluoride was not determined because of the thermal instability of the fluoride and the lack of proper reference compounds against which physical properties of the unknown might be compared. Experiments were run which showed that the type and rate of mixing over a wide range of conditions had little effect on the final product composition. However, this reaction is heterogeneous with both a hydrocarbon and hydrofluoric acid phase present; for this reason, an exact treatment is difficult. I t has been reported that alkylation of m-xylene with I-decene in the presence of hydrofluoric acid gives 2-(2,4dimethylpheny1)decane ( 8 ) . Analysis of the product from the hydrofluoric acid alkylation of m-xylene with 1decene by gas-liquid chromatography showed the presence of four major components (Table IV) . Infrared analysis of the entire sample showed it was mainly a 1,2,4-trisubstituted alkylbenzene. These results suggest the product is mainly a mixture of 2-, 3-, 4-,and 5(2,4-dimethylphenyl)decane. Similar results were obtained in the hydrofluoric acid alkylation of toluene with l-dodecene. A mixture of four major isomers, identified by infrared analysis as mainly para disubstituted benzenes, was found. These facts plus the relative amounts present indicated the mixture
Table IV. These Are the Products of Hydrofluoric Acid Alkylation of Certain Aromatics Column Relative Retention Producta Tim? Wt. % m-XYLENE WITH ~ - D E C E N E 2- (2,4-Dimethylphenyl) decane 2.5 20 3-(2,4-Dimethylphenyl) decane 2.0 24 4-(2,4-Dimethylphenyl)decane 1.8 20 5-(2,4-Dimethylphenyl) decane 1.7 29 Not identified 1.4 3 Not identi5ed 1.3 4 Reference n-hexadecane 1.0
-
-
too-+-q
I
w
z a
was composed of 2-, 3-, 4-, 5-, and 6-(4-methylphenyl)dodecane(Table IV). Sulfuric Acid Alkylation. Alkylation of benzene with 1-dodecene in the presence of sulfuric acid gave a mixture of secondary phenyldodecanes. While it has been stated (2) that isomer distribution is independent of catalyst strength in the range 91 to loo%, some work in our laboratory suggests that it may not be independent of the amount of catalyst present. That is, with less sulfuric acid present less isomerization may occur. Treating pure 2- or 6-phenyldodecane with benzene and sulfuric acid did not result in isomerization. The 2- or 6phenyldodecane was recovered unchanged. These results are similar to those observed in this work with hydrofluoric acid and with other information in the literature on the sulfuric acid isomerization of phenylalkanes (4). Aluminum Chloride Alkylation. The aluminum chloride alkylation of benzene with 1-dodecene at 4 5 ” to 53’ C. gave the same mixture of isomeric phenyldodecanes obtained by treating either 2- or 6-phenyldodecane with aluminum chloride and benzene under alkylation conditions. In the alkylation with this catalyst 1-dodecene disappears very rapidly. Analysis of aliquots of a reaction showed that the initial composition of the phenyldodecanes formed
\
MOLE 2-PHENYLDODECANE BENZENE ALUMINUM CHLORIDE WATER TEMPERATURE 50°C
80-
V
W
RATIO
I .o 32 0.087 0.0035
a
0
f3
60-
>z w
PHENYLDODECANE
I
CL
*.
TOLUENE WITH 1-DODECENE 2-(4-Methylphenyl) dodecane 3.8 13 Unknownb 3.4 3 3-(4-Methylphenyl) dodecane 3.1 15 4-(4-Methylphenyl) dodecane 2.7 17 5- and 6-(4-Methylphenyl)dodecane 2.5 44 Unknownb 2.2 a Reference n-hexadecane 1.0 a Tentative identification. Possibly 1,2dialkylbenzenes.
a
E
20
-
0
0
IO
20
30
40
50
60
TIME, MINUTES 2-Phenyldodecane isomerizes to an equilibrium mixture of phenyldodecanes in the presence of benzene and aluminum chloride VOL. 52. NO. 10
OCTOBER 1960
835
changed with time. The first aliquot had 4070 2-phenyldodecane. The concentration of 2-phenyldodecane decreased with time to a n equilibrium value of 32% with compensating increases in the amount of 4-, 5-, and 6-phenyldodecanes. Within experimental error the amount of the 3- isomer remained constant. Although the temperature effect was not completely explored, aluminum chloride alkylation at 30” C. gave the same distribution of isomers and dialkylbenzene found at higher temperatures. Very low temperature (0’ to 10’ C.) alkylation was not investigated. Treating 2-phenyldodecane with benzene and aluminum chloride decreased the amount of the 2- isomer and formed the 3-, 4-,5, and 6- isomers. The first aliquot of this reaction showed the presence of all the other secondary phenylalkanes in significant amounts. In these experiments where aliquots were analyzed over a period of time, the procedure was to remove a small aliquot, quench it in water, wash it free of the catalyst, and analyze it directly by gas-liquid chromatography. I n order to obtain a material balance in these reactions, a known amount or a nonreactive material such as n-decane was added initially to serve as an internal standard in the gas-liquid chromatographic analyses. By knowing the initial ratio of this internal standard to starting material it was possible to calculate the composition of the aliquots by relating the measured peaks to the standard and the original ratio. I n this way it was established that very little disproportionation occurred. Discussion The first step in the hydrofluoric acid alkylations is probably the formation of a solvated intermediate between alkene and catalyst. +H+
RCHzCH
=
CH2
7 -H+
RCH,CH
1-
CH,
.1 H+
1
This intermediate can lead to the reversible formation of 2-fluorododecane which by repeated elimination and addition of a molecule of hydrofluoric acid could account for the isomerization. r
I
1
Alternatively. the isomerization could occur by hydride ion shifts under the influence of hydrofluoric acid or hydrofluoric acid plus benzene.
[ [
RCH*CHCH,]
t + 1
C8HG 2-phenyldodecane
--f
hvdride shift
lXy€!CH&]
% 3-phenyl-dodecane
If this path is correct then the observed appearance and disappearance of alkyl fluoride during the course of the reaction may only be a reversible side reaction which is not important to the formation of phenylalkane. The mechanism for the sulfuric acid catalyzed alkylation is probably similar to that for hydrofluoric acid. Addition of a molecule of sulfuric acid to I-alkene by a carbonium ion process gives alkyl hydrogen sulfate. Isomerization may then result from repeated elimination and addition of sulfuric acid. Alkylation of the intermediate carbonium ions would give the mixture of secondary phenylalkanes that are found in the product. Alternatively, the isomerization may occur by hydride ion shifts involving an intermediate catalyst-alkenearomatic complex. According to one commonly accepted mechanism, the first step in the alkylation is the formation of an intermediate pi complex between catalyst, alkene, and aromatic (12). This collapses to form a sigma complex in which the alkyl group is now attached to the aromatic nucleus. Elimination of a proton from this sigma complex, perhaps through another pi complex, produces the alkylated benzene. For the aluminum chloride catalyzed reactions all of these steps are reversible. Isomerization may occur during or just preceding a trans-alkylation of the alkyl group to another aromatic nucleus from the original pi complex between catalyst, alkene, and aromatic. This mechanism has been suggested for the isomerization of phenylpentanes in the presence of this catalyst ( 4 ) . The large amount of 2-phenyldodecane found in the product during the early part of the reaction may be accounted for if the rate of alkylarion is sufficiently favored over that of isomerization and trnns-alkylation. With time the equilibrium mixture of phenyldodecanes is formed through reverse alkylation steps and isomerization from the catalyst1 C,H,
I -
2-phenyldodecane
RCHZCHCHS
I
dodecane
F -H+
836
INDUSTRIAL AND ENGINEERING CHEMISTRY
alkene-aromatic complex. Once the isomerization has begun it is probably very fast. It may also not be necessary to build up a highconcentrationof 3-phenyldodecane before any of the 4- isomer can form. As pointed out previously in the aluminum chloride isomerizaLion of 2phenyldodecane, the decrease in the concentration of this isomer occurs simultaneously with an increase in the concentrations of all of the other isomers. The initial alkene-catalyst-aromatic pi complex may also be the intermediate from which the isomerization occurs in the hydrofluoric and sulfuric acid systems. With these catalysts one or more of the general steps for alkylation enumerated earlier are not reversible, at least not to any significant extent. This would account for the failure of the pure isomers to isomerize. I n the presence of hydrofluoric or sulfuric acids, the rate of the alkene isomerization must be sufficiently rapid compared io the rates of the alkylation steps to account for the mixtures of isomers found in rhe products. Acknowledgment The author is indebted to his many collegues who assisted him in this hvork and especially C. F. Spencer and L. P. Lindeman for gas-liquid chromatography and high mass spectrometer analyses. He also wishes to express appreciation to Oronite Chemical C o . for its interest and funds to support the investigation. literature Cited (1) Asahara, T., Takaji, Y., J. Chem. Soc. Japan; Ind. Chem. Sect. 58, 147 (1955). (2) Asinger, F., Geiseler, G., Beetz, W., Chem. Ber. 92, 755 (1959). (3) Baumgartner, F. N., ISD. ENG. CHEM. 46, 1349 (1954). (4) Burwell, R. L.. Jr., Shields, A . D., J . Am. Chem. SOC.77, 2766 (1955). (5) Ferris, S. W., “Handbook of Hydrocarbons,” Academic Press, New York, 1955. ( 6 ) Gray, F. W., Gerecht, J . F . , Krems, I. J., J . Org. Chem. 20, 51 1 (1955). (7) Griess, W., Fette, Seifen, Anstrichmittel 57, 24, 168, 236 ( 1 955). (8) Hunt, M. ( t o E. I. du Pont de iYemours & Co., Inc.), U. S. Patent 2,467,131
(April 12, 1949).
( 9 ) James, A. T., Martin, A . J. Biochem. J . 50, 679 (1952). (10) Lenneman, W. L., Hites, K. Komarewsky, V. I., J . o r g . Chem. 463 (1954). (11) Meyerson, S.: Appl. Spectroscopy
P., U.,
19, 9,
120 (1955). (12) Nelson, K. LeRoi, Brown, H. C., “Chemistry of Petroleum Hydrocarbons,” Vol. 111, Chap. 56, Reinhold, iSew York, 1955. (13) Ray, N. H., .I. Appl. Chem. 4, 22 (1954). (14) Tjepkema, J. J., Paulis, B., Huijser, H. W., Proc. Fifth World Petrol. Congr., Sect. IV, Paper 21, New York, 1959. RECEIVED for review March 11, 1960 ACCEPTED June 23, 1960 Division of Petroleum Chemistry, 136th Meeting, ACS, Atlantic City, h-. .J.> September 1959.